User authentication has become very important along with the development of information and communication technology. In recent years, various user authentication methods have been developed in addition to the classic authentication method of using a code such as a password or the like.
In particular, user authentication methods using biometric information recognition are free from the risk of forgetting or losing the code such as the password or the like and are thus characterized by ease of use and reliable security. The user authentication methods using biometric information have been mainly used in embedded systems such as personal computers (PC) or automated teller machines (ATM). Recently, advancements in technology have expanded the use of the user authentication methods using biometric information to even mobile devices.
Among the various types of biometric information, fingerprints maintain the same pattern throughout the life term. Even if a fingerprint is damaged by external factors, it retains its original form through healing, and the probability of having an identical fingerprint with another is as low as one in one billion. Due to such advantages, the fingerprint is a common form of biometric information that is adopted as a means for authentication, and the field of fingerprint authentication is also very diverse.
In using fingerprint authentication, if the quality of the fingerprint image, which is acquired through a fingerprint input window that allows the user to input biometric information, is not guaranteed at a certain level or higher, an authentication failure may occur. In order to solve this problem, one of the user authentication methods for acquiring authentication images utilizes a background image acquired through the fingerprint input window. The background image is an image acquired through the fingerprint input window in a state where a biometric feature (e.g., the fingerprint of a user's finger) is not in contact with the fingerprint input window. By determining a difference between the fingerprint image and the background image, it is possible to reduce the adverse effect caused by residual fingerprints on the fingerprint input window.
However, in the case of an in-display fingerprint sensor, a large number of residual fingerprints remain on the fingerprint input window because users are frequently placing their fingerprint on the fingerprint input window during user authentication and touch input events. Therefore, it is difficult to obtain a high-quality image of the fingerprint due to the adverse effect caused by the residual fingerprints. This may result in a decrease in authentication performance.
Patent document 1: Korean Patent No. 10-1432011 (published on Aug. 13, 2014)